基于二部图匹配的资源共享方法研究--以阿里云物联网云平台为例

Research on resource sharing method based on bipartite graph matching--taking Aliyun Internet ofThings cloud Platform as an example

物联网的发展已经深入人们的生活,为用户带来便利的生活体验。但与此同时,随着智能设备的发展,物联网用户的大量增加加深了内容和资源的分配矛盾。因此,研究在分布式资源分配中引入了图理论与匹配理论,提出了基于二部图与匹配理论的三维稳定匹配算法。该算法利用匹配理论思想对用户的资源进行高效分配。在对比实验中,该算法在CU数量最多时,也可在1000次内达到收敛,与对比算法相比具有最好的收敛性。其ASDP最高为80.2%,ADT最短为2 s。其用户的满意度,即内容请求接受比例和用户服务质量(Quality of Service,QoS)需求满足率,最高可达91.0%和96.5%。可见,该算法具有最佳的性能表现,为物联网用户资源共享问题提供了新的方案。

The development of the Internet of Things has penetrated into people's lives and brought convenient life experience to users.But at the same time,with the development of intelligent devices,the large number of Internet of Things users has deepened the contradiction between the distribution of content and resources.Therefore,graph theory and matching theory are introduced into distributed resource allocation,and a three-dimensional stable matching algorithm based on bipartite graph and matching theory is proposed.The algorithm uses the idea of matching theory to efficiently allocate users'resources.In the comparison experiment,the al-gorithm can converge within 1000 times when the number of CU is the largest,and has the best convergence compared with the com-parison algorithm.The maximum ASDP is 80.2%,and the minimum ADT is 2s.Its user satisfaction,that is,the content request ac-ceptance ratio and the user quality of service(QoS)demand satisfaction rate,can reach 91.0%and 96.5%respectively.It can be seen that the algorithm has the best performance and provides a new solution for the problem of user resource sharing in the Internet of Things.